An interestedness timing estimation device that estimates a timing related to interestedness of a user toward content is known. The interestedness timing estimation device acquires usage history that includes usage times for content subject to the estimation of the timing related to the interestedness of the user and for content other than the estimation subject. Then, the interestedness timing estimation device computes usage ratios of the content subject to estimation with respect to content other than the estimation subject from the acquired usage history during respective times. Then, the interestedness timing estimation device estimates a timing related to interestedness of the user toward the content subject to estimation based on the computed usage ratios during the respective times.
Further, a server device that determines a degree of interest toward a webpage provided from a website over the Internet is known. This server device acquires, from a terminal device, scroll operation information for a display region of the terminal device and content specifying information for specifying content displayed in the display region. Then, based on the acquired scroll operation information, the server device computes the degree of interest of the user toward the content specified by the acquired content specifying information.
Further, an interestedness estimation device that estimates a level of interestedness toward respective documents using multiple grades is known. The interestedness estimation device calculates a normalized viewing time for each document viewed by a user, this being a per-character viewing time for the respective viewed document, and generates data expressing a relationship between the normalized viewing times and a corresponding number of viewed documents. The interestedness estimation device then integrates the relationship data over time and normalizes the integrated relationship to update information expressing a per-user relationship between normalized viewing time and an interestedness score. Next, the interestedness estimation device references the information expressing the updated relationship between normalized viewing time and interestedness score to estimate a score indicating a level of interestedness of the user toward a document viewed by the user.